Molecular bases of hexokinase deficiency

Ž .Biochimica et Biophysica Acta 1360 1997 211–221

Molecular bases of hexokinase deficiency

Marzia Bianchi, Rita Crinelli, Giordano Serafini, Camilla Giammarini, Mauro Magnani )

Institute of Biological Chemistry ‘G. Fornaini’, UniÕersity of Urbino, Via Saffi 2, 61029-Urbino, Italy

Received 29 October 1996; accepted 17 December 1996

Abstract

Ž .Hexokinase ATP: D-hexose 6-phosphotransferase, EC 2.7.1.1; HK deficiency is a rare disease where the predominantclinical effect is nonspherocytic hemolytic anemia. We have previously shown that the only patient for which hexokinasedeficiency has been so far investigated at molecular level is a double heterozygote carrying a T1667 ™C substitution on one

Ž .HK type I allele and a 96 bp deletion concerning nucleotides 577 to 672 in the HK cDNA sequence in the other allele. Toinvestigate whether these mutations found in the patient with the hexokinase variant referred to as ‘HK-Melzo’ could beassociated with hexokinase deficiency, we have expressed in E. coli the wild-type human hexokinase type I and twodifferent mutants carrying the T™C nucleotide substitution at position 1667 and the nt 577–672 deletion, respectively.Wild-type human recombinant hexokinase is expressed in bacterial cells as a soluble catalytically active enzyme that, uponpurification to homogeneity, exhibited the same kinetic properties of human placenta hexokinase type I. Both mutanthexokinases were also expressed as soluble recombinant proteins under the same conditions, but they showed an impairedcatalytic activity with respect to the wild-type enzyme. In particular, the T1667 ™C substitution, causing the amino acidchange from Leu529 to Ser, is responsible for the complete loss of the hexokinase catalytic activity, while the 96 bp deletioncauses a drastic reduction of the hexokinase activity. Taken together, both mutations explain the hexokinase deficiencyfound in the patient with the ‘HK-Melzo’ variant. q1997 Elsevier Science B.V. All rights reserved.

Ž .Keywords: Hexokinase deficiency; Site-directed mutagenesis; Human

1. Introduction

ŽHexokinase in mammalian cells ATP: D-hexose.6-phosphotransferase, EC 2.7.1.1; HK is the first

enzyme of the glycolytic pathway that commits glu-cose to catabolism by catalyzing its phosphorylation

Ž .to glucose 6-phosphate Glc 6-P with MgPATP asphosphate donor and is considered one of the rate-

w xlimiting enzymes of glycolysis 1 . Four differentisozymes named type I, II, III and IV or glucokinase,

) Corresponding author. Fax: q39 722 320188.

which vary in their tissue distribution, kinetic andmolecular properties are present in mammalian tis-

w xsues 2 . Among all these, HK type I is the predomi-nant glucose phosphorylating activity in those tissueswhich share a strict dependence upon glucose utiliza-tion for their physiological functions such as braintissue, erythrocytes, platelets, lymphocytes and fi-

w xbroblasts 3–6 . Hexokinase deficiency is a rare dis-ease where the predominant clinical effect is non-spherocytic hemolytic anemia. The reasons why thisgenetically determined molecular lesion has only thisconsequence at clinical level are still unknown. Only14 cases of hexokinase deficiency have been de-

0925-4439r97r$17.00 Copyright q 1997 Elsevier Science B.V. All rights reserved.Ž .PII S0925-4439 96 00080-4

( )M. Bianchi et al.rBiochimica et Biophysica Acta 1360 1997 211–221212

w xscribed so far 7 , two of which have been studied inw xour laboratory 8–10 : in one case, designated ‘HK-

Melzo’, the HK type I cDNA from the patient withhemolytic anemia was also characterized at molecularlevel. PCR amplification and sequence of the entire‘HK-Melzo’ cDNA revealed the presence of a dele-tion and of a single nucleotide substitution, both in

w xheterozygous form 11 . We thus concluded that thepatient carried a point mutation at nt 1667 on one HK

Ž 529allele which causes the amino acid substitution Leu.to Ser and a 96 bp deletion on the other allele and

was therefore unable to produce any wild-type fullycatalytically active hexokinase type I. To investigatewhether both the nucleotide substitution at position1667 and the 96 bp deletion were responsible for thehexokinase deficiency in the patient, we have nowgenerated two hexokinase mutants carrying the pointmutation and the 96 bp deletion, respectively. Ex-pression of wild-type recombinant human hexokinaseand the two mutants confirmed that both mutationsimpaired, although to a different extent, the hexoki-nase catalytic activity.

2. Materials and methods

2.1. Materials

ŽlhHEX-15 clone was from ATCC American Type.Culture Collection, MD . pET expression vectors and

Ž .E. coli strain BL21 DE3 were purchased from No-Ž .vagen Madison, WI . ‘Sculptore in vitro mutagene-

sis system’ and M13mp19 RF DNA were fromAmersham as well as ‘Sequenase Version 2.0 DNA

Ž . w 35 xSequencing Kit’ USB and a S dATP used forDNA sequence analysis. Mutagenic primers andPhenyl Sepharose CL-4B were obtained from Phar-macia LKB Biotechnology. PVDF membranes used

for solid phase assay and Goat anti rabbit IgGhorseradish peroxidase conjugate were purchasedfrom Bio Rad.

2.2. Synthesis of ‘HK-Melzo’ cDNA

The cDNA of ‘HK-Melzo’ was prepared by re-verse transcription of mRNA purified from a lympho-blastic cell line derived from the lymphocytes of the

w xpatient, as detailed in 11 .

2.3. Construction of wild-type human hexokinase typeI expression Õector

The construction of the expression vector forwild-type human hexokinase type I was performedessentially as recently described by Fromm and

w xcoworkers 12 for the human brain hexokinase, withslight modifications. The cDNA sequence for expres-sion of wild-type HK I was obtained from the lamb-

Ž .dahHEX-15 clone purchased from ATCC, MD . Thisclone includes nucleotides 38-3598, encoding a com-plete 917 amino acid protein, and moreover contains43 bp 5X untranslated and 765 bp 3X untranslated

w xsequence 13 . The 316 bp EcoRI fragment oflhHEX-15 containing the amino terminal 91 codonsof the HK I open reading frame was ligated into the

- Ž .Bluescript KS plasmid Stratagene previously cutŽ y .with EcoRI. The resultant plasmid KS -316 bp HK

was submitted to the oligonucleotide directed site-specific mutagenesis procedure to create a NdeIrecognition site at the ATG start codon of the HK Iopen reading frame, exactly as described by Liu, F. et

w xal. 12 . The 272 bp NdeI- EcoRI fragment obtainedby digestion of the mutant plasmid was gel-purifiedand then ligated into the pET-3a expression vectorŽ .Novagen . The 3245 bp fragment, containing nt354-3598 of HK type I cDNA, was isolated from the

Ž . Ž . Ž .Fig. 1. Construction of pET-3a HK a and pET-3a HKm1667 b, c expression vectors. a To construct the pET-3a HK expressionvector the following steps were performed: the pET-3a plasmid was digested with NdeI and EcoRI and then ligated with the 272 bp

Ž . Ž .NdeI-EcoRI fragment of HK I cDNA containing nt 82-353 . The resultant pET-3a HK 82-353 plasmid was linearized with EcoRI andŽ .subsequently ligated with the 3245 bp EcoRI fragment of HK I cDNA concerning nt 354-3598 to obtain the pET-3a HK expression

Ž . Ž 529 .vector. b To construct the mutant Leu to Ser expression vector the 4302 bp XbaI-PstI fragment, containing the entire cDNA wascut from pET-3a HK and subcloned into the M13mp19 to perform the site-directed mutagenesis T1667 ™C using the mutagenic primer

Ž . 1667HK 1656M. c To obtain the final construct named pET-3a HKm1667 the 2904 bp NcoI-PstI fragment containing the desired T ™Csubstitution was cut from the M13 mutant RF DNA and then recloned into the plasmid pET-3a HK from which the wild-type NcoI-PstIfragment had been removed. Further details of these procedures are reported in Section 2.

( )M. Bianchi et al.rBiochimica et Biophysica Acta 1360 1997 211–221 213


same lhHEX-15 clone by digestion with EcoRI,followed by gel purification. This 3245 bp EcoRIfragment, comprising the coding region for aminoacid residues 92-917 as well as the 3X untranslated

region, was then ligated into the pET-3a plasmid thatalready contained nt 82-353 of the HK N-terminalsequence. The maintenance of the correct orientationof the two partial HK I cDNA fragments was verified


by nucleotide sequence analysis. The expression vec-tor with the full-length HK type I cDNA was namedpET-3a HK and the scheme used for its constructionis shown and detailed in Fig. 1a.

( 529 )2.4. Construction of the mutant Leu to Ser hex-okinase type I expression Õector

Site-directed mutagenesis at nt 1667 was per-formed using the ‘Sculptore in vitro mutagenesissystem’ from Amersham, according to the instruc-tions of the manufacturers and essentially as de-

w xscribed by Zeng, C. and Fromm, H.J. 14 . Briefly,the expression plasmid pET-3a HK was digested withthe restriction enzymes XbaI and PstI. A 4302 bpXbaI-PstI fragment, containing the entire HK cDNAinsert plus flanking regions of the vector, was sub-cloned into the polylinker region of M13mp19 at theXbaI and Pst I sites and was used as a template tomake the site-directed change. For the T™C substi-tution at nt 1667 the 24-mer long mutagenic primer,designated HK 1656M, was 5X-TGGTGACTTCTCG-GCCCTGGATCT-3X. The single strand DNA fromM13 transformants was sequenced in order to iden-tify the desired T1667 ™C substitution and also toexclude the presence of random mutations in theentire portion of the open reading frame submitted to

Ž .the mutagenesis protocol Fig. 1b . DNA sequencingwas carried out by dideoxy chain termination method

w xof Sanger et al. 15 , using Sequenase Version 2.0. Inorder to transfer the mutant to the expression plasmid

Ž .pET-3a Fig. 1c a replicative form DNA of M13mutant was cut at two unique restriction sites withNcoI and Pst I. The 2904 bp NcoI-PstI fragment wasgel-purified and then recloned into the plasmid pET-3a HK from which the wild-type NcoI-PstI fragmenthad been removed. The desired T™C mutation inthe expression plasmid termed pET-3a HKm1667was reconfirmed by DNA sequence analysis.

2.5. Construction of the 96 bp deleted hexokinasetype I expression Õector

To express in E. coli a mutant recombinant hexok-inase containing the nt 577-672 deletion, the follow-

Ž .ing steps were performed: 1 a PCR amplification of‘HK-Melzo’ cDNA which gave two PCR products in

Žthe same ratio: the first one was 697 bp long from nt.390 to nt 1087 and the second one was 601 bp long

because it contained the 96 bp deletion. The coupleof primers used for this amplification were: HK 3Ž . X Xsense 5 -GAATGTTCACATGGAGTCCGAG-3

Ž . Xand HK 6 antisense 3 -GAGTGGGCTCCCT-X Ž .TCAAATTGT-5 . 2 The 601 bp fragment was gel-

Žpurified, ligated into the TA cloning vector Invitro-.gen and then subjected to DNA sequence analysis.

Ž .3 The cloned deleted insert was cut with Bcl Iwhich cleaves at bases 462 and 764 and the resultant206 bp long fragment was ligated with pET-3dNH HK plasmid from which the wild-type 302 bp2

Ž .Bcl I undeleted fragment had been removed Fig. 2a .Ž .4 The expression plasmid obtained designated pET-3d NH HKdel was digested at two unique restriction2

sites with XbaI and NcoI: the 1302 bp fragmentobtained, containing the nt 577–672 deletion, wasthen recloned into the pET-3a HK expression vectorfrom which the corresponding wild-type undeleted

Ž . Ž .fragment 1398 bp long had been removed Fig. 2b .The 96 bp deletion in the final expression plasmidnamed pET-3a HKdel was verified by DNA sequenceanalysis.

2.6. Expression of wild-type and mutant hexokinaseproteins

The recombinant plasmids pET-3a HK, pET-3aHKm1667 and pET-3a HKdel were transformed into

Ž .the E. coli strain BL21 DE3 to overexpress thewild-type and mutant hexokinase proteins. The induc-

Ž .Fig. 2. Construction of the pET-3a HKdel expression vector. a The 601 bp deleted PCR product obtained by amplifying the ‘HK-Melzo’cDNA with the couple of primers HK 3-HK 6 was inserted into the TA cloning plasmid. This cloned deleted insert was then cut withBcl I to obtain a 206 bp long fragment containing the nt 577–672 deletion. This fragment was ligated with pET-3d NH HK vector,2

Ž .previously digested with Bcl I to obtain the plasmid termed pET-3d NH HKdel. b pET-3d NH HKdel was cut with XbaI and NcoI:2 2

the 1302 bp fragment, containing the 96 bp deletion, was then recloned into the pET-3a HK expression vector, from which theŽ .corresponding wild-type undeleted fragment 1398 bp long had been removed. The final expression vector was named pET-3a HKdel.

Further details of these procedures are reported in Section 2.


tion experiments were initially performed accordingto the instructions of the manufacturers. The opti-mization of the recombinant human hexokinase ex-pression in prokaryotic cells using the pET system aswell as the selected temperature of induction are

w xdescribed in 16 . Also the preparation of bacterialextracts for hexokinase activity and protein assays

w xwas as reported in 16 .

2.7. Purification of recombinant wild-type humanhexokinase type I

The purification of recombinant wild-type humanhexokinase type I was performed as described previ-

w xously 16 with some modifications. Briefly, afterammonium sulfate fractionation the pellet was resus-

Žpended in 10 ml of HIC buffer 5 mM sodiumpotassium phosphate buffer, pH 7.55, containing 5mM glucose, 3 mM 2-mercaptoethanol and 30%Ž . .wrv ammonium sulfate and submitted to Hy-drophobic Interaction Chromatography, using a

Ž .Phenyl Sepharose CL-4B column 1.8=13 cm . Elu-tion was performed with a linear gradient of ammo-

Žnium sulfate from 30 to 0% 500 ml in each cham-.ber . Fractions containing hexokinase activity were

collected, dialyzed, concentrated and directly loadedonto a Blue-A column. The dye-ligand chromatogra-

w xphy step was performed as previously reported 16 ,with the exception that washing and elution bufferscontained 90 mM KCl instead of 25 mM.

2.8. Hexokinase assay

Hexokinase activity was measured at 378C spec-trophotometrically in a system coupled with glucose

Ž .6-phosphate dehydrogenase G6PD; EC 1.1.1.49 or,in the Glc 6-P inhibition studies, in a coupled enzyme

Ž .system with pyruvate kinase PK; EC 2.7.1.40 andŽ .lactate dehydrogenase LDH; EC 1.1.1.28 as de-

w xscribed in 17 . One unit of hexokinase activity isdefined as the amount of enzyme which catalyzes theformation of 1 mmol of Glc 6-P or ADPrmin at378C. Kinetic studies were performed in 100 mM

w xTris-HCl, pH 7.2 17 .

2.9. Determination of protein

Protein concentration was determined using thew xBradford method 18 with bovine serum albumin as

standard, or spectrophotometrically by measuring theabsorbance of solutions at 280 nm against appropriateblanks.

2.10. SDS-PAGE and immunoblotting

SDS-PAGE of the expressed wild-type and mutanthuman HK I was performed in 14% polyacrylamide

w xgels according to the method of Laemmli 19 , whileWestern blot analysis was carried out exactly as

w xdescribed in 20 .

2.11. Solid phase assay

Bacterial extracts in SDS sample buffer were incu-bated at 1008C for 5 min, then diluted at the finalconcentration of 0.4 mgrml in 150 mM NaCl. Asample of human wild-type recombinant hexokinase,

Ž .purified to homogeneity 150 Urmg protein , wastreated as above with the exception that the finalconcentration was 0.04 mgrml. Appropriate furtherdilutions were made with the same buffer containing

Ž .0.02% wrv SDS so that the signal obtained afterstaining for hexokinase was linear with the amount ofsample applied and was within the range of hexoki-

Žnase amounts used as standard 0.062–0.015.mgrwell . Fifty microliters volume for samples and

hexokinase standards were loaded onto PVDF mem-Ž .brane 0.2 mm with the aid of a 96-well Minifold

Ž .apparatus Bio Rad . After 2 hours at room tempera-Žture the wells were rinsed with TBS 50 mM Tris-Cl

.pH 7.5 and 150 mM NaCl and drawn to neardryness. The filter was then removed, blotted dry,and fixed. Fixed blots were then immunochemicallystained for hexokinase using an immunoaffinity-iso-

Žlated rabbit anti-human hexokinase antibody ob-.tained as in Ref. 20 and goat anti-rabbit peroxidase-

conjugate IgG as a secondary antibody. The detectionsystem was the enhanced chemiluminescenceŽ .Amersham . Quantitation of recombinant hexokinasepresent in bacterial extracts was by densitometry at492 nm of the resulting film. Absolute values wereobtained by comparison with a standard curve gener-ated by plotting the desitometric values of knownamounts of hexokinase standard applied to the samefilter.


3. Results

3.1. Expression and characterization of wild-typehuman recombinant hexokinase I

The expression of native recombinant human hex-okinase in transformed bacterial culture was inducedby the addition of IPTG at a final concentration of0.4 mM. After induction an aliquot of bacterial cellswas used for preparation of bacterial extracts as

w xdescribed in 16 . 20 mg of soluble cellular proteinswere separated by SDS-PAGE and stained withCoomassie Blue. Induction with IPTG resulted in theappearance of a prominent protein of about 100 kDawhich migrated at the same position as purified hu-man placenta hexokinase type I; it was not present in

Ž Ž .the induced control cells i.e. E. coli BL21 DE3cells transformed with the plasmid pET-3a lacking

.the HK cDNA insert . Western blot analysis con-firmed that the 100 kDa expressed protein was indeedhuman hexokinase since it reacted with a monospe-cific rabbit anti-human hexokinase type I antibodyŽ .data not shown . The highest specific activity valuewas obtained by reduction of the induction tempera-

w xture from 378C to 228C for 15 h as detailed in 16 . InŽ .these selected conditions E. coli cells BL21 DE3

transformed with the expression vector pET-3a HKcontained a biologically active hexokinase with aspecific activity of 2.00"0.26 Urmg of protein,corresponding to a concentration of 13.2 " 2.7HKUrml, whereas induced control cells had a hex-okinase specific activity of only 0.088"0.02 Urmgof protein, corresponding to a concentration of 0.43"0.05 HKUrml of bacterial lysate. The wild-typehuman recombinant hexokinase was purified to ho-mogeneity as described in Section 2. Table 1 summa-rizes the purification steps and indicates the average

Ž . Ž .Fig. 3. SDS-PAGE A and Immunochemical detection B ofwild-type, Leu529 to Ser mutant and 96 bp deleted recombinanthexokinases. 20 mgs of soluble post-induction cell proteins from

Ž .bacterial cells transformed with pET-3a HK plasmid lane 1 ,Ž . ŽpET-3a HKm1667 plasmid lane 2 , pET-3a HKdel plasmid lane

. Ž .3 and pET-3a plasmid alone lane 4 were submitted to SDS-Ž .PAGE and Coomassie staining A or electroblotted as described

w x Ž . Žin 20 B . Position of molecular weight markers with corre-.sponding M in kDa are indicated on the left. The blot wasr

probed with an immunoaffinity-isolated rabbit anti-human hexok-inase type I IgG and immunoreactive bands were visualized byhorseradish-peroxidase-labelled goat anti-rabbit IgG antibodies,followed by ECL chemiluminescence detection.

yield from approx 10 preparations. The homogeneousenzyme was further characterized for its kinetic andregulatory properties in parallel with hexokinase type

Table 1Purification of wild-type human hexokinase type I expressed in E. coli

Ž . Ž . Ž . Ž .Purification steps Total act. U Spec. act. Urmg Recovery % Purification fold

Cell-free extract 1250 2.0 100 1.0Ž .Ammonium sulfate 70% of DEAE peak 725 11.0 58 5.5

Hydrophobic interaction chromatography 475 42.0 38 21.0Blue A dye-ligand chromatography 325 150 26 75

All values are means of approx. 10 different experiments.


Table 2Kinetic properties of recombinant wild-type human hexokinase and placental hexokinase type I

Property Wild-type recombinant Placental hexokinasehexokinase type I type I

Ž .K mM for MgPATP 1.40"0.20 0.90"0.10mŽ .K mM for glucose 64.5"6.0 60.0"5.8mŽ .K mM for Glc 6-P 28.5"3.0 30.0"4.0iŽ .K mM for Glc 1,6-P 54.0"5.1 47.0"5.0i 2

Removal of Glc 6-P and Yes YesGlc 1,6-P inhibition by 2.0 mM P2 i

Ž .The K for MgPATP was determined at saturating glucose concentration 5 mM , while the K for glucose was derived at saturatingm mŽ . Ž .MgPATP concentration 5 mM . The K values were determined with MgPATP as the varied substrate 0.5–2.0 mM at 5 mM-glucosei

concentration by a replot of slopes, obtained from double reciprocal plots, versus inhibitor concentrations. The patterns all indicatedcompetitive inhibition. All the experiments were performed in 100 mM Tris-HCl, pH 7.2, at 378C. The removal of Glc 6-P and Glc 1,6-P2

w xinhibition by P was evaluated as in 17 . The values are means"S.D. of three experiments.i

w xI purified from human placenta 21 . The apparentK for MgPATP of the wild-type recombinant en-m

zyme was 1.4 mM and that of the placental enzyme0.9 mM. Both enzymes are inhibited competitivelywith respect to MgPATP by glucose 6-phosphateŽ . Ž .Glc 6-P and glucose 1,6-bisphosphate Glc 1,6-P2

Žwith removal of the inhibition by phosphate Table.2 . Inhibition of wild-type recombinant HK by these

two phosphorylated sugars was non-competitive withŽ .respect to glucose results not shown .

3.2. Expression of the recombinant human hexoki-( 529 )nase mutant Leu to Ser and of the 96 bp deleted

mutant

pET-3a HKm1667 expression vector and pET-3aŽ .HKdel plasmid were transformed into BL21 DE3 E.

coli strain. The induction procedure was as described

for wild-type human HK I as was the preparation ofbacterial extracts used for hexokinase activity andprotein assays. The specific activities of hexokinase

Ž .in extracts from induced BL21 DE3 cells trans-formed with pET-3a HKm1667 and pET-3a HKdelare shown in Table 3. The comparison of mutant andwild-type hexokinase specific activities in bacterialextracts suggests that while the site-directed changeT1667 ™C was responsible for the complete loss of

Žhexokinase catalytic activity, the 96 bp deletion cor-.responding to amino acids 166-197 of HK I caused a

drastic reduction in the activity. Attempts to obtainsome measurable hexokinase activity in pET-3aHKm1667 transformed cells by changes in tempera-ture of induction, duration of induction and celldensity before induction provided values similar tothose reported in Table 3. SDS-PAGE and Westernblotting experiments using induced bacterial extracts

Table 3Ž .Hexokinase specific activities in extracts from induced BL21 DE3 cells, transformed with different expression plasmids

Ž .BL21 DE3 transformed cells HK spec. act. HK spec. act.Ž . Ž .Urmg of bacterial protein Urmg of immunoreactive protein

Induced pET-3a 0.088"0.02 rInduced pET-3a HK 2.00 "0.26 150"7.5Induced pET-3a HKm1667 0.095"0.02 3.5"0.4Induced pET-3a HKdel 0.350"0.06 12.3"0.7

w xThe determinations of hexokinase specific activities were performed on bacterial lysates prepared as described in 16 . The amount of HKimmunoreactive protein in cell lysates was obtained by a solid phase immunoassay as reported under Section 2. All values aremeans"S.D. of five experiments.


confirmed that indeed both recombinant proteins areŽexpressed in soluble form in the bacterial cells Fig.

.3 . However, since different recombinant proteins canbe expressed at different levels in E. coli cellsandror the efficiency of induction with IPTG couldvary from one experiment to another the expressionof the specific activity of hexokinase as units per mgof total protein could be inappropriate. For this rea-

Žson, the activity of the recombinant hexokinases ex-pressed as the difference between total and endoge-

.nous hexokinase activity was normalized to im-munoreactive hexokinase protein as determined bythe solid phase immunoassay. This method allows fordirect comparison between wild-type and mutant re-combinant hexokinases and also for the exclusion ofthe interfering endogenous enzyme which is not rec-

Ž .ognized by the antibody used not shown . The spe-cific activity values given as units of HK per mg of

Ž .immunoreactive protein Table 3 clearly indicatethat, while the Leu529 to Ser mutation completelyabolished the HK catalytic activity, the 96 bp deletiondrastically reduced the activity of the enzyme.

4. Discussion

The only clinical manifestation of hexokinase defi-w xciency appears to be hemolytic anemia 22 . The

reasons why impairment of an important step inglucose utilization has only this consequence at theclinical level are still unknown: one possibility is thatthe HK reaction is catalyzed by four isozymes, asstated in the Section 1, and therefore only tissuesstrongly dependent on only one isozyme could beaffected by such a mutation. Hexokinase in the ery-throcytes is of central importance because it is one ofthe rate-limiting steps in anaerobic glycolysis, theonly pathway by which ATP is produced in thesecells. Only 14 cases of hexokinase deficiency have

w xbeen described so far 7 and the only patient forwhich the enzymatic defect of HK type I has beeninvestigated at the molecular level is a double het-erozygote carrying a T™C substitution at nt 1667 onone HK type I allele and a 96 bp deletion on the

w xother allele 11 . To investigate whether the twomutations found in the ‘HK-Melzo’ cDNA were re-ally responsible for the hexokinase deficiency, we

have now expressed in E. coli the two mutant hexok-1667 Žinases carrying the T ™C substitution which

529 .causes the amino acid change from Leu to Serand the 96 bp deletion, respectively, and then com-pared their properties with those of the wild-typehuman recombinant hexokinase type I. The wild-typehuman recombinant HK is expressed in bacterial cellsas a soluble catalytically active protein that, uponpurification to homogeneity, shows the same kineticand regulatory properties of human placenta HK type

w xI 21 . Both mutant hexokinases were also expressedas soluble proteins under the same conditions and thehexokinase specific activity was measured in thesoluble bacterial extracts. The results obtained showthat while the 96 bp deleted hexokinase is onlypartially active, the Leu529 to Ser mutant is com-pletely inactive. Thus, we conclude that the hexoki-nase catalytic activity is abolished when the non-polaramino acid residue Leu529 is replaced by the polarSer and is strongly affected when the 96 bp deletion,corresponding to the amino acid residues 166–197 inthe HK type I, is introduced. Since the three-dimen-sional structure of HK type I and the organization ofits gene are not known, we can only refer to otherevolutionarily correlated enzymes such as yeast hex-

w x w xokinase 23 and human glucokinase 24 for whichthe gene structure has been reported. Due to the highdegree of homology of the three hexokinases men-tioned above, we have tentatively placed the 96 bpdeletion within amino acids 162 to 193 of human

w xglucokinase, which include all exon 6 24 , suggest-ing that it may occur by aberrant splicing. The 96 bpdeletion removes a portion of the N-terminal half ofhexokinase type I, comprising the amino acid residuesfrom 166 to 197. Since previous results from our

w xlaboratory 20 and considerable evidence in the liter-w xature 14,25,26 suggest that the catalytic site of HK

type I is located in the C-terminal domain of theenzyme and that the N-terminal half is devoid of anycatalytic activity, the low activity found in the mutantdeleted HK was unexpected. However, although thekinetic and regulatory properties of the isolated C-terminal half of HK are very similar, if not identical,to those of the native enzyme, a role for the N-termi-nal region of hexokinase I in maintaining the properconformation of the full-length HK cannot be ex-cluded. In previous studies on the ‘HK-Melzo’ vari-ant, we found that the residual enzyme activity in the


Fig. 4. Comparison of partial hexokinase and glucokinase aminoŽ .acid sequences see also Ref. 1 . Sequences compared are: b1,

bovine type I; h1, human type I; m1, murine type I; I, rat type I;II, rat type II; III, rat type III; hIV, human type IV; IV, rat typeIV; sm, hexokinase from Schistosoma mansoni; ygk, yeast glu-

Žcokinase; yst, yeast hexokinase isozymes A and B the sequenceshown is that of the A isozyme, with residues that differ in the B

.isozyme shown immediately below . The arrow marks the Leuresidue which is conserved in all these hexokinases so far se-

Žquenced this Leu residue corresponds to the amino acid 529 in.the human HK type I .

erythrocytes of the patient was about 20% and there-fore it could be argued that the 96 bp deletion in theN-terminal half of the enzyme could generate animproperly folded hexokinase protein with a reducedcatalytic activity. As regards the Leu529 ™Ser muta-tion, this change corresponds to a substitution of anon-polar amino acid residue with a polar one. More-over, this Leu529 residue is highly conserved in all

w xthe hexokinases 1 as reported in Fig. 4, whichshows the striking similarities in sequence betweenall the isozymes of the hexokinase family. This Leu529

is placed within the b-sheet of the small domain ofw xyeast hexokinase 23,27–29 which is known to un-

dergo major conformational changes during substratebinding. In particular, the conformational change fromthe ‘open’ to the ‘closed’ form, induced by glucose isconsidered to be a necessary step for catalysis andthus it can be supposed that even the substitution ofonly one residue may impair the enzyme catalyticfunction. Additional crystallographic studies will cer-tainly be necessary to further support these conclu-sions. However, our results provide direct evidencethat the amino acid substitution Leu529 ™Ser and the96 bp deletion found in human hexokinase deficiencyare responsible for the impairment of catalytic activ-ity found in the only case in which a hexokinasedefect has been investigated at molecular level so far.

Acknowledgements

This work was supported in part by funding fromC.N.R. and M.U.R.S.T.

References

w x Ž .1 Wilson, J.E. 1995 Rev. Physiol. Biochem. Pharmacol. 126,65–198.

w x Ž .2 Katzen, H.M. and Schimke, R.T. 1965 Proc. Natl. Acad.Sci. U.S.A. 54, 1218–1225.

w x Ž .3 Shows, T.B. 1974 Cytogenet. Cell Genet. 13, 143–145.w x Ž .4 Chern, C.J. 1976 Cytogenet. Cell Genet. 17, 338–342.w x5 Rijksen, G., Staal, G.E.J., Beks, P.J., Streefkerk, M. and

Ž .Akkerman, J.W.N. 1983 Biochim. Biophys. Acta 719,431–437.

w x6 Magnani, M., Stocchi, V., Rossi, L., Dacha, M. and For-`Ž .naini, G. 1984 Bull. Mol. Biol. Med. 9, 9–15.

w x Ž .7 Miwa, S. 1983 Am. J. Haematol. 14, 381–391.w x8 Magnani, M., Stocchi, V., Cucchiarini, L., Novelli, G.,

Ž .Lodi, S., Isa, L. and Fornaini, G. 1985 Blood 66, 690–697.w x9 Magnani, M., Chiarantini, L., Stocchi, V., Dacha, M. and`

Ž .Fornaini, G. 1986 J. Inher. Metab. Dis. 9, 129–139.w x10 Magnani, M., Stocchi, V., Canestrari, F., Dacha, M., Balestri,`

P., Farnetani, M.A., Giorgi, D., Fois, A. and Fornaini, G.Ž .1985 Br. J. Haematol. 61, 41–50.

w x Ž .11 Bianchi, M. and Magnani, M. 1995 Blood Cells Mol. Dis.21, 2–8.

w x Ž .12 Liu, F., Dong, Q., Myers, A.M. and Fromm, H.J. 1991Biochem. Biophys. Res. Commun. 177, 305–311.

w x Ž .13 Nishi, S., Seino, S. and Bell, G.I. 1988 Biochem. Biophys.Res. Commun. 157, 937–943.

w x Ž .14 Zeng, C. and Fromm, H.J. 1995 J. Biol. Chem. 270,10509–10513.

w x Ž .15 Sanger, F., Nicklen, S. and Coulson, A.R. 1977 Proc. Natl.Acad. Sci. U.S.A. 74, 5463–5467.

w x Ž .16 Bianchi, M., Serafini, G., Corsi, D. and Magnani, M. 1996Protein Expression Purif. 7, 58–66.

w x17 Magnani, M., Stocchi, V., Dacha, M. and Fornaini, G.`Ž .1984 Biochim. Biophys. Acta 804, 145–153.

w x Ž .18 Bradford, M.M. 1976 Anal. Biochem. 72, 248–254.w x Ž .19 Laemmli, U.K. 1970 Nature 227, 680–685.w x20 Magnani, M., Bianchi, M., Casabianca, A., Stocchi, V.,

Ž .Daniele, A., Altruda, F., Ferrone, M. and Silengo, L. 1992Biochem. J. 285, 193–199.

w x21 Magnani, M., Stocchi, V., Serafini, G., Chiarantini, L. andŽ .Fornaini, G. 1988 Arch. Biochem. Biophys. 260, 388–399.

w x Ž .22 Valentine, W.N. and Paglia, D.E. 1984 Blood 64, 583–591.w x Ž .23 Harrison, R.W. 1985 Crystallographic refinement of two

isozymes of yeast hexokinase and the relationship of struc-ture to function. Ph.D. dissertation, New Haven, CT, YaleUniversity.

w x24 Stoffel, M., Froguel, PH., Takeda, J., Zouali, H., Vionnet,N., Nishi, S., Weber, I.T., Harrison, R.W., Pilkis, S.J.,


Lesage, S., Vaxillaire, M., Velho, G., Sun, F., Iris, F., Passa,Ž .PH., Cohen, D. and Bell, G.I. 1992 Proc. Natl. Acad. Sci.

U.S.A. 89, 7698–7702.w x Ž .25 White, T.K. and Wilson, J.E. 1989 Arch. Biochem. Bio-

phys. 274, 375–393.w x Ž .26 Arora, K.K., Filburn, C.R. and Pedersen, P.L. 1993 J.

Biol. Chem. 268, 18259–18266.

w x Ž .27 Anderson, C.M., McDonald, R.C. and Steitz, T.A. 1978 J.Mol. Biol. 123, 1–13.

w x Ž .28 Anderson, C.M., Stenkamp, R.E. and Steitz, T.A. 1978 J.Mol. Biol. 123, 15–33.

w x Ž .29 Bennett, W.S. Jr. and Steitz, T.A. 1980 J. Mol. Biol. 140,211–230.

Molecular bases of hexokinase deficiency

Documents

Transcript of Molecular bases of hexokinase deficiency